There is presently not much evidence that chemotherapy is of benefit in malignant CNS tumors. The principal reason for this may be the existence of the blood brain barrier in the case of metastatic tumors whereas with regard to primary CNS tumors, drug sensitivity may be more important than the drug delivery problem. The presence of this barrier prevents most presently available chemotherapeutic agents from penetrating the CNS parenchyma and cerebrospinal fluid. This proposal is designed to try to overcome the problem of drug delivery with both animal and clinical studies. Hypertonic solutions have been shown by a number of investigators to disrupt transiently the blood brain barrier and has enabled agents such as microperoxidase, penicillin, Evan's blue labeled albumin, and various neuro-transmitters which normally do not permeate the blood brain barrier to enter the central nervous system. Previous studies in this laboratory have shown that we can achieve a 10 to 50 fold increase in CNS methotrexate levels in the rat and dog after intra-arterial mannitol blood brain barrier disruption. No permanent neurologic defects have been noted. Close observation of the animals for up to two weeks following reversible barrier disruption has not revealed any neurologic complication. Neuropathologic examinations of rat brains after one day, seven days, and two weeks has not shown any evidence of histologic changes. Nonetheless, damage to the eye has been reported in some animals. Following barrier disruption C.T. scans of dogs which received intravenous conray 60 and intravenous Evan's blue before osmotic barrier disruption, show that there are marked areas of enhancement on the C.T. scan which correspond exactly to the areas of Evan's blue-albumin penetration of the blood brain barrier. There is a 20 C.T. density unit difference between the enhanced and unenhanced areas of the brain. The C.T. scan thereby is a sensitive non-invasive monitor of both the extent and the degree of blood brain barrier disruption. The present studies are planned to evaluate the effects of blood brain barrier disruption in man. We plan to determine the clinical and laboratory criteria documenting the presence of barrier disruption as well as the toxicity of barrier disruption. We then will evaluate the access of methotrexate or other drugs to malignant CNS tumors and the clinical responsiveness of such a chemotherapeutic maneuver.